Section 2: Data Analysis
Introduction
Previous studies have shown that heavy-handed regulations on the housing market will reduce housing availability and increase cost, but some may argue that Maine’s market is intrinsically different. Maine is a national outlier in many ways, having the highest median age of any state and being the most forested state by percentage of land. Maine is also known as “Vacationland” and a small-town state, with far more of its population living in small to moderate towns than large cities, like many other states.
Due to Maine’s many differences from other states, there will inevitably be objections to applying nationwide findings to Maine. After all, if Maine is so different from other states, then its housing market may behave differently as well. Thus, this section studies the correlation between the status of land use zoning in towns and housing prices, and a separate analysis of local minimum lot size requirements.
Using a 2022 report on land use and zoning by the Mercatus Center, Maine Policy, in conjunction with Professor James Siodla of Colby College, compares the housing markets of various towns throughout Maine. By focusing on the relationship between restrictive local land use policies and high demand and growth in the housing market, we can see if there is a correlation between the two variables. Once a correlation is established, we can analyze the form of the relationship.
In other sections of this report, we have emphasized policies such as rent control and inclusive zoning, but these policies are not widespread in Maine outside of a few larger cities, such as Portland. As of 2021, out of New England’s almost 200 localities with inclusive zoning, only 1% were in Maine. Meanwhile, Maine has about 9% of New England’s population, showing that for New England at least, Maine has a disproportionately low amount of inclusive zoning. Rent control policies were similarly distributed, with only larger cities and cities in southern Maine appearing to have experimented with it.
Because of the limited Maine-specific data on these two policies, we will focus on land use regulations used more widely throughout the state. Both land use zoning and minimum lot size are policies used commonly enough in Maine to have accessible data and also show enough variance to isolate the town-by-town effects of the policies. Many Maine towns are restrictive in one of these categories and nonrestrictive in the other, so it should be possible to identify whether a relationship exists between these policies and housing market growth.
While many modern cities employ heavy-handed land use regulations regardless of size, some towns throughout Maine do not employ use-based zoning. Use-based zoning restricts the type of buildings that can be built and the things they can be used for based on where they are located within the municipality. While many towns in Maine do use zoning, around 200 do not employ it at all.
Many of these towns are small towns in rural northern Maine with very inactive housing markets; however, this is not true for all of them. Many of these towns are medium-sized and located in central, midcoast or even southern Maine, including Lebanon, Paris, Harpswell, and Monmouth. These towns all have a population of 5,000 or more, are either west or south of Augusta, and do not have the traditional form of land use-based zoning.
Comparing the housing markets of similarly sized and located zoned and unzoned towns will not be a perfect answer as to the impact of land use regulations, and even unzoned towns can have strict land use regulations in other categories. However, a possible correlation between zoning and inflated housing prices can still provide helpful insight into how Maine’s housing market functions.
Minimum lot size is another important kind of land use regulation. Additionally, the more rural a town is, the more it tends to restrict minimum lot size. Thus, finding a correlation between stricter minimum lot size and higher housing costs will show that this is not simply “city housing is more expensive,” but that the housing regulations are genuinely impacting prices. Minimum lot size, in short, disallows people from developing or using properties for specific purposes without the lot being a minimum size. By restricting minimum lot size, towns thus limit the density at which people can build, the number of people able to live in their locality, and the type of residential properties that can be developed.
Zoning vs Nonzoning
Using the data collected by the Mercatus Center in their “Regulating without Zoning in Maine Towns” report, we have organized Maine towns by zoned and unzoned status. Then, by cross referencing that with Geographic Information Survey (GIS) data and regional Zillow home price indexes, we can detect any correlations between towns’ zoned or unzoned status and their local home price. If either category’s home price is significantly higher, that may mean the cost of housing, and thus the housing market, is impacted by whether a town uses land use zoning. Other potential explanations exist for a difference in home prices between these categories, which we address later in our research.
When looking at the Maine-specific data, towns with land-use zoning codes experience higher average house prices. According to our data, this significant difference in price between zoned and unzoned towns has existed since at least 2010. This difference appeared at its highest at 10.8% higher home prices in zoned cities in 2010. However, the difference seems to be increasing again recently, from 5.9% in 2020 to 6.9% in 2023.
Because zoned towns tend to be larger and more urban than unzoned towns, they also permit denser housing unit designs, such as multi-family housing. However, a troubling trend has emerged since 2020, where zoned towns have permitted 9.06 fewer total units, 1.95 more single-family units, and 11.01 fewer multi-family units than unzoned towns. Thus, despite the demand for Maine housing increasing, many towns in Maine have reduced their housing production rather than increasing it to meet demand. Between 2010 and 2024, Maine’s population grew by over 70,000 people, however, we have not had housing growth at the same rate. Additionally, the gap between the permits issued in zoned and unzoned cities has been growing, suggesting that the price difference is partially caused by a change in supply.
This change is also concernedly recent, as from 2010-2019, there was no statistically significant difference between the two categories of housing stock growth. This indicates that the zoning-based difference is very recent, possibly due to the increased demand from new Mainers moving here from other parts of New England. As the demand for housing increases, unzoned towns increase supply to respond to demand, while zoned towns do not respond to that shift in demand in the same way.
The second concern to control for is mistaking the source of price changes, which can be caused by differences in demand or differences in supply, all else equal. One market’s price may be lower for several reasons, and while one reason may be a regulatory burden, another possible reason is a difference in local demand. Certain small northern towns have seen a decline in population over the past decade, while significant population increases occurred in southern Maine. Of course, housing is inherently likely to face higher demand in places where people want to live. This will naturally increase the price of high-demand towns, meaning that one other reason zoned towns are pricier to live in is that more people may simply want to live there.
Accounting for this was not as straightforward as accounting for town size, but we combined the data from Mercatus and Zillow with towns’ permitting requests, which helped us account for demand and supply shifts. Markets with higher demand will naturally also receive more building permit requests. So, by considering this data, we were able to contextualize the above differential between zoned and unzoned towns, further proving that the regulatory burden is to blame for the increased costs instead of an increase in local demand for housing.
The above graph is an illustration that one might find in an economics classroom displaying a leftward supply shift. Economists use these graphs to simulate the market forces of demand and supply in a simplified way. The “Y” axis is always shown as price, and the “X” axis is displayed as quantity. Because people demand products less when they are cheaper, the demand curve slopes downward, and similarly, more significant quantities of goods are supplied by firms when they sell for higher prices. By shifting the supply curve left, we see that the equilibrium price of homes increases and housing quantity decreases, precisely what we have seen in Maine.
Regulations on the housing market lead to a leftward shift of the supply curve, as do many regulations that restrict suppliers of a good or service. Regulations on housing typically create some extra cost or burden on suppliers, which means their profits from constructing housing are reduced. When earnings from building housing are reduced, housing suppliers are willing to build fewer houses, which causes a leftward shift of the supply curve. This leftward shift increases the average market price and reduces the total quantity of housing available.
Zoned towns in Maine have been shown to have less affordable housing markets, negatively impacting housing accessibility. Assuming that the disparity between these two binary categories indicates a more significant negative relationship between zoning and housing market growth, towns may benefit from even smaller steps toward deregulation. Even if such steps are minor, the speculative long-term nature of the housing market may encourage developers to pick one town over its more restrictive neighbors.
Zoning is just one of the local land use policies in Maine we analyzed, and it would be incorrect to assume that this inverse relationship between land use regulation and market growth is limited to zoning policy alone. In the next section, we will analyze the impact of restrictive minimum lot sizes across nonzoned towns in Maine.
Minimum Lot Size
While we have now shown a correlation between inflated housing prices and the presence of zoning, some may feel this is limited to only one kind of land use regulation. Providing evidence of a relationship between minimum lot size and housing price should further solidify the case that land use regulation in Maine harms the housing market. If our theory is correct, the more restrictive the local land use policy is, the greater the burden on the housing market. While zoned towns have a wide divergence in minimum lot size from one zone to another, the hundreds of unzoned towns in Maine typically only employ one minimum lot size or occasionally two. Focusing on these towns will allow us to understand the relationship between this policy and the local affordability of houses.
For several reasons, the minimum lot size is more challenging to analyze than the zoning and nonzoning status. First, because zoned towns employ so many diverse zoning methods with separate minimum lot sizes, it would require far more data and more significant resources to include them in this section entirely. Additionally, many unzoned towns have low populations, making the divergence of lot sizes and home prices very volatile. To account for these factors, we have explicitly focused on unzoned Maine towns with a population greater than 2,000, of which there are 44.
The minimum lot size of unzoned towns can diverge greatly, with some cities, such as Lebanon, Bowdoin, and Pittston having a minimum lot size of two acres. These towns appear unzoned because they desire a small town “leave me alone” atmosphere, but this can motivate them to create incredibly restrictive minimum lot sizes to prevent denser property uses from obstructing the towns’ aesthetics.
Meanwhile, other towns appear unzoned to allow residents to do what they want with their properties. This motivation impacts not only the local zoning policy but also the minimum lot size, and many unzoned towns in Maine, such as Madison, Jay, or Blue Hill, have no minimum lot size. There are two caveats to this section, which complicates our analysis somewhat, but the negative relationship between minimum lot size and housing affordability was still observed.
The first problem we encountered was shoreland zoning. Maine has a statewide law requiring stricter zoning policies for properties directly abutting shoreland to preserve the aesthetic and environmental values of Maine’s shoreland. This law applies to some but not all of these towns. The good news is that even coastal towns such as Harpswell are not universally affected because the shoreland zoning restrictions only reach about 200 feet from shore. Thus, unless a property directly touches the waterfront, it is unlikely to be directly affected.
The second problem was a greater challenge. Maine also has a statewide minimum lot size requirement of 20,000 square feet. However, this only applies to properties that use private subsurface waste disposal systems, such as septic systems, so properties with public sewer access or no sewer are allowed to be as small as the owners want. Additionally, there are still two reasons that this complicates the minimum lot size analysis.
First, some towns with minimum lot sizes of less than 20,000 square feet or no minimum lot size also have little to no public sewer access. This makes it almost impossible to judge whether a property in the town is impacted by the statewide mandate, at least not without going door to door to every property in town to see whether they have private subsurface waste disposal systems.
This combines with the second issue, which is that the towns with public sewer systems for only part of the town, in effect, have an invisible “reduced minimum lot size” zone of which not even the town itself knows the full coverage. Understanding the coverage of this zone is crucial because it impacts whether a property has a 20,000-square-foot minimum lot size, a smaller minimum, or no minimum lot size. Again, all of the towns with partial coverage do not know how many residential properties in their town have public sewer access. Aside from being an obstacle to our research, this is incredibly concerning because it means that the local land use boards who decide minimum lot size for their town also have no idea what percentage of the town population they are impacting and how.
The justification for the statewide legal requirement of 20,000 square feet for properties with private subsurface sewage disposal is understandable. Human waste can have significant negative impacts on the local environment when condensed into denser areas. However, a situational minimum lot size burdens many prospective developers by complicating land use requirements. This policy creates an extra land use requirement that invisibly applies only to specific properties and frequently is not referenced in the local land use code. The conflict may even confuse some developers who may be unsure whether the local land use code takes precedence over the state’s, and may need to hire legal counsel to explain that the state code takes precedence.
We were able to account for this statewide requirement in a somewhat indirect way. While a map of each town’s sewer system could not be easily found, we could locate some towns’ total sewer connection numbers. Furthermore, this included differentiation between a commercial, industrial, and residential sewer/water district connection. By cross-referencing this data with the number of households in the town, we estimated the number of households that did not have sewer access. Knowing this allows us to assess the proportion of the town impacted by the statewide minimum lot size requirement.
This had potential risks, such as those inherent to combining data from multiple sources, as one source’s definition of a household may be more limited or expansive than another. Additionally, some sewer connections may have been mislabelled as residential, or over or undercounted. Regardless, this data was used to estimate the percentage of towns’ households without public sewers rather than generate a precise number. Since general trends were being analyzed rather than individual towns, the impact that individual, incorrectly labeled lots will have is somewhat minimal.
While we could not discern which specific parts of each town were impacted by the statewide minimum lot size statute, we could find a general estimate of the percentage of each city affected by the law. By accounting for this in our data, we could still compare the rate of the towns that genuinely had the minimum lot size mandated by the town rather than the state. Since many towns had a larger minimum lot size regardless of sewer access, this consideration did not impact those towns.
While the above factors should be considered in adding context to our analysis, as should the smaller sample size of the towns we examined, the difference between towns with and without minimum lot restrictions was significant. It is even more remarkable than the difference between zoned and unzoned towns. In 2010, towns with minimum lot sizes had a 40.9% higher home price than those without. This number decreased in 2020 to 20.1%, then 34.7% in 2023. While this gap is smaller than in 2010, it still shows that towns without minimum lot sizes have significantly lower housing costs than those with minimum lot sizes.
When analyzing the minimum lot size, it should be noted that the sample size was much smaller. This is because zoned towns almost universally have minimum lot sizes that vary throughout the towns’ zones, making it essentially impossible to compare them. However, focusing on medium-sized unzoned Maine towns still provides enough data to create a somewhat reliable outline of the relationship between minimum lot size and housing price. Emphasizing these towns in our analysis allows us to avoid the effects that being in a big city can have on local housing prices. Overall, a 10,000-square-foot increase in minimum lot size was associated with a 4% increase in average house price.
While this difference may not seem significant, it should be noted that even 4% of a half-a-million-dollar house is $20,000. In 2024, the average home price in Maine rose above $400,000, with 4% of that figure totaling $16,000. If this 4% increase represents a causal relationship, Maine could reduce minimum lot sizes statewide by 10,000 square feet and save the average home buyer $16,000. Additionally, this figure does not include the interest that most homeowners would be paying on that amount when it is incorporated into their mortgage.
Furthermore, several towns have highly restrictive minimum lot sizes over 80,000 square feet. If this correlation represents a causal relationship, then those towns’ minimum lot sizes may increase the cost of a house locally by around 32% compared to if that town had no minimum lot size at all. One such town is Lebanon, which has a minimum lot size of two acres for any lot with a building on it.
Houses in Lebanon regularly sell for north of $400,000. If the 4% increase per 10,000 square feet remains stable, Lebanon’s minimum lot size ordinance inflates the average home’s price by about 34.8%. That works out so that the mean home sale price would be closer to $297,000 if Lebanon had no minimum lot size. If our analysis holds, Lebanon’s minimum lot size costs the average home buyer over $100,000.
Conclusion
While Maine goes against national trends in many ways, it is clear that this does not apply to the fundamental relationship between land use regulations and the housing market. As with any market, rules and fees are passed on to consumers. Natural supply shortages are created by reducing the number of developers who can afford to develop new housing and the number of buyers who will pay inflated prices. This is caused by inclusive zoning, rent control, and similar rental-focused policies. In Maine, however, zoning and minimum lot size restrictions mainly reduce housing availability.
Due to Maine’s “small town” nature, these two policies can disproportionately affect us compared to other states. Large cities tend to be more flexible on minimum lot size and zoning requirements than small towns, but far less of Maine’s population lives in cities than in other New England states. Vermont is the state that has the smallest share of its population living in urban centers, only 35%, but Maine is a close second with 39% urban population. The next most urban state is West Virginia, with a distant 45% urban population. Because the predominant number of lower-income Mainers live in smaller, rural towns, we can expect this to burden them disproportionately.
Both minimum lot sizes and zoning reduce the number of houses that can be built in a town. Not surprisingly, reducing minimum lot sizes correlates positively with housing availability and more affordable housing options. This has already been established elsewhere in New England. Still, we now know that the same relationship also exists in Maine housing markets in particular. Similarly, reducing zoning restrictions allows for a larger percentage of the town to be developed into housing, reducing the cost of the average home. This relationship was also shown to exist nationally, but even in Maine-specific markets, this effect exists.
The evidence shows a strong correlation between housing costs and regulations, best explained by a leftward supply shift. While there is some expected difference in demand between zoned and unzoned towns, this should not be understood to be the sole cause of this difference and does not account for the discrepancy in permit requests. Even in Maine specifically, the evidence shows a clear decline in housing availability when strict local zoning or minimum lot size ordinances are present.
This is not to say that other policies in Maine do not have similar or even worse effects. The literature and direct evidence show that housing market regulations can be divided into two groups: those that reduce the flexibility of available housing options and those that reduce the profitability of creating and providing housing. Both minimum lot size and zoning requirements fall primarily into the first category, as they restrict the amount and type of housing that can be provided. Meanwhile, profitability restrictions such as inclusive zoning or rent control are less widespread in Maine, making comparative Maine-centric data challenging to find.
Maine-specific evidence shows that these more restrictive land use regulations reduce availability and affordability. In a period when much of the country and Maine faces a severe housing crisis, Maine localities must employ land use policies that synergize with and understand market forces rather than work against them. In the next section, we will rely on the findings discussed in sections one and two to draw conclusions on which policies would benefit Maine’s housing market most. However, it is essential to understand from this section that not only do housing market restrictions negatively impact housing market health in the abstract, but they have also been shown to place unnecessary burdens on home buyers and developers here in Maine.